发现单细胞大模型注意力机制主要反映基因共表达，而非独特调控信号。
Systematic Evaluation of Single-Cell Foundation Model Interpretability Reveals Attention Captures Co-Expression Rather Than Unique Regulatory Signal
- 构建37项分析框架，系统评估单细胞大模型可解释性。
- 注意力模式虽有结构，但预测扰动效果不如简单基因基线（AUROC 0.81-0.88）。
- 适用于关注模型可解释性与调控网络推断的研究者。
我们提出一个系统评估框架——包含37项分析、153次统计检验、四种细胞类型及两种扰动方式——用于评估单细胞基础模型的机制可解释性。将该框架应用于scGPT和Geneformer，发现注意力模式具有分层组织的结构化生物信息：早期层对应蛋白互作，晚期层对应转录调控，但这种结构对扰动预测无增量价值：基因级基线表现优于注意力与相关边（AUROC 0.81–0.88 对比 0.70），成对边得分贡献为零，且因果消融调控头不造成性能下降。该发现从K562细胞推广至RPE1细胞；注意力与相关性的关系具有上下文依赖性，但基因级主导性具有普适性。针对注意力特定的缩放失败，提出细胞状态分层可解释性（CSSI），提升调控网络恢复达1.85倍。该框架为领域建立可复用的质量控制标准。
原文摘要 · Abstract (English)
We present a systematic evaluation framework - thirty-seven analyses, 153 statistical tests, four cell types, two perturbation modalities - for assessing mechanistic interpretability in single-cell foundation models. Applying this framework to scGPT and Geneformer, we find that attention patterns encode structured biological information with layer-specific organisation - protein-protein interactions in early layers, transcriptional regulation in late layers - but this structure provides no incremental value for perturbation prediction: trivial gene-level baselines outperform both attention and correlation edges (AUROC 0.81-0.88 versus 0.70), pairwise edge scores add zero predictive contribution, and causal ablation of regulatory heads produces no degradation. These findings generalise from K562 to RPE1 cells; the attention-correlation relationship is context-dependent, but gene-level dominance is universal. Cell-State Stratified Interpretability (CSSI) addresses an attention-specific scaling failure, improving GRN recovery up to 1.85x. The framework establishes reusable quality-control standards for the field.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。